Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Sun Baked

macrumors G5
Original poster
May 19, 2002
14,945
165
Originally posted by M.Isobe:
_Power dissipation and on-die L2_

IBM posted interesting presentation about the power dissipation of PPC750FX/GX.
+ Because SRAM needs a power to maintain the content, the processor with large on-die L2 has a large static power dissipation component,
+ 750FX has a large static power dissipation component and the frequency has less effect on power dissipation,
+ 750GX's power dissipation is dominated by static component and the frequency has a minor effect on power dissipation,
+ IBM uses RC5-72 to measure max. power dissipation.
Due to sub-threshold leakage, the static power dissipation becomes more serious in 90nm processors. So, small on-die L2 of 970FX seems to be a design decision to lower the power dissipation.
NOTE: That IBM says Power Supply and Cooling System must be adequate for Maximum Power Dissipation -- we've been seeing a lot of "Typical" numbers.

With these numbers being half of Maximum (at least for the IBM processors we're using), our hopes of specific processors becoming mobile mean Maximum Power Dissipation is quite important.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.